Doctoral dissertation (article) (G5)

Development of Secure Software : Rationale, Standards and Practices




List of AuthorsRindell Kalle

PublisherTurku Centre for Computer Science

PlaceTurku

Publication year2019

eISBN978-952-12-3832-1

URLhttp://urn.fi/URN:ISBN:978-952-12-3832-1

Self-archived copy’s web addresshttp://urn.fi/URN:ISBN:978-952-12-3832-1


Abstract

The society is run by software. Electronic processing of personal and financial data forms the core of nearly all societal and economic activities, and concerns every aspect of life. Software systems are used to store, transfer and process this vital data. The systems are further interfaced by other systems, forming complex networks of data stores and processing entities.This data requires protection from misuse, whether accidental or intentional. Elaborate and extensive security mechanisms are built around the protected information assets. These mechanisms cover every aspect of security, from physical surroundings and people to data classification schemes, access control, identity management, and various forms of encryption. 
Despite the extensive information security effort, repeated security incidents keep compromising our financial assets, intellectual property, and privacy. In addition to the direct and indirect cost, they erode the trust in the very foundation of information security: availability, integrity, and confidentiality of our data. Lawmakers at various national and international levels have reacted by creating a growing body of regulation to establish a baseline for information security. Increased awareness of information security issues has led to extend this regulation to one of the core issues in secure data processing: security of the software itself. Information security contains many aspects. It is generally classified into organizational security, infrastructure security, and application security. Within application security, the various security engineering processes and techniques utilized at development time form the discipline of software security engineering. The aim of these security activities is to address the software-induced risk toward the organization, reduce the security incidents and thereby lower the lifetime cost of the software. Software security engineering manages the software risk by implementing various security controls right into the software, and by providing security assurance for the existence of these controls by verification and validation. A software development process has typically several objectives, of which security may form only a part. When security is not expressly prioritized, the development organizations have a tendency to direct their resources to the primary requirements. While producing short-term cost and time savings, the increased software risk, induced by a lack of security and assurance engineering, will have to be mitigated by other means. In addition to increasing the lifetime cost of software, unmitigated or even unidentified risk has an increased chance of being exploited and cause other software issues. 
This dissertation concerns security engineering in agile software development. The aim of the research is to find ways to produce secure software through the introduction of security engineering into the agile software development processes. Security engineering processes are derived from extant literature, industry practices, and several national and international standards. The standardized requirements for software security are traced to their origins in the late 1960s, and the alignment of the software engineering and security engineering objectives followed from their original challenges to the current agile software development methods. The research provides direct solutions to the formation of security objectives in software development, and to the methods used to achieve them. It also identifies and addresses several issues and challenges found in the integration of these activities into the development processes, providing directly applicable and clearly stated solutions for practical security engineering problems. 
The research found the practices and principles promoted by agile and lean software development methods to be compatible with many security engineering activities. Automated, tool-based processes and the drive for efficiency and improved software quality were found to directly support the security engineering techniques and objectives. Several new ways to integrate software engineering into agile software development processes were identified. Ways to integrate security assurance into the development process were also found, in the form of security documentation, analyses, and reviews. Assurance artifacts can be used to improve software design and enhance quality assurance. In contrast, detached security engineering processes may create security assurance that serves only purposes external to the software processes. The results provide direct benefits to all software stakeholders, from the developers and customers to the end users. 
Security awareness is the key to more secure software. Awareness creates a demand for security, and the demand gives software developers the concrete objectives and the rationale for the security work. This also creates a demand for new security tools, processes and controls to improve the efficiency and effectiveness of software security engineering. At first, this demand is created by increased security regulation. The main pressure for change will emanate from the people and organizations utilizing the software: security is a mandatory requirement, and software must provide it. This dissertation addresses these new challenges. Software security continues to gain importance, prompting for new solutions and research.


Last updated on 2021-24-06 at 09:23